What are the pros and cons of:

• Series Cells vs
• Parallel Cells + a Boost converter?

Thoughts:

• It seems both would provide the same voltage and current.
• Except the parallel boost would have some losses.
• However, if the series pack also has a boost (in order to maintain steady voltage for the load), then both circuits would have converter losses.

Series + Boost benefits:

• Since the series pack voltage is higher than parallel, the series boost doesn't need to boost voltage as far as the parallel boost. So the series boost would have lower losses than the parallel boost.

Parallel + Boost benefits:

• passive balancing
• if cells discharge at different rates such that some cells reach discharge cutoff before other cells, the fully-discharged cells can be removed from the pack without interruption or change in voltage of the boost output.

 

What is the Voltage that You need ?
What are the Voltages of your Batteries ?
.
.
.

 

IR losses tend to be greater whenever current increases, as does the cost of conductors. A weak cell will degrade the performance in either configuration. Balancing is a trade off between longevity and demand. Supervisory circuits can inform you of what your state of charge is, and how efficiently it can be delivered, from the cells perspective.

if cells discharge at different rates such that some cells reach discharge cutoff before other cells, the fully-discharged cells can be removed from the pack without interruption or change in voltage of the boost output.

How could parallel cells have different voltages

 

There is not enough information provided to decide which scheme is preferred. The question would be justhow much current is required and for how much time, as well as what sort of battery.

 

Other considerations to include in your excellent bullet points is the cost and size constraints.

Having said this… One cannot make a blanket statement without values, at least voltage in and out, and minimum/ maximum current load.
Here Excel is your best friend.

 

How could parallel cells have different voltages

Even if cells begin at full and balanced, if there's variation in health, age, resistance, capacity, etc (even if same cell type), and if the discharge rate is faster than the balance rate (meaning, if the cells are discharging faster than they can maintain balance), then i think the cells might not reach fully discharged voltage at the same time.

One cannot make a blanket statement without values, at least voltage in and out, and minimum/ maximum current load.

The assumption not stated in my question, because i thought obvious, is we're using same number of cells in each arrangement, and that the outputs are equivalent.

For example, 10 Li cells charged to 4V ea in parallel, with a 50V boost would theoretically have same voltage and current as 10 Li cells charged to 4V ea in series, with a 50V boost

So i think this question doesn't require knowing "what current and voltage do you need" to answer the question.

 

The questions posed in post #2 still stand. Which scheme is best still depends on the application. Different applications require different solutions.
If, as an example, the application requires a rather high current for a fairly short time, it may work best to use a series string of cells to provide the required power without any boost system that would be less than 100% efficient. I have serviced such a system on a few occasions.
Usually, if a suitable voltage can be provided without any conversion required, that will be the most efficient scheme, if the current requirement is within the capability of a reasonable quantity of parallel series strings.
Consider that If "one size fits all", it will probably not fit anyone perfectly.

 

Series + buck is probably better.

Even if cells begin at full and balanced, if there's variation in health, age, resistance, capacity, etc (even if same cell type), and if the discharge rate is faster than the balance rate (meaning, if the cells are discharging faster than they can maintain balance)

If one cell discharges faster, its voltage drops below that of the other cells and it stops discharging until the others catch up. This process is continuous, and thus the terminal voltages remain the same. If one cell was at a lower voltage, current would actually flow into that cell.

Since all battery terminals are connected together, they, by definition, at the same voltage.

A weak cell will simply supply less than its share of the current throughout the discharge cycle. It is even OK to parallel cells of different caoacacities. .

 

In nicad strings I have seen one series cell destroyed by being reverse charged. Inlarge lead-acid strings I have seen one battery blown up by load current on the series string..

 

There's no "cheap-way" to bend the rules.

If You try to use just what ever Batteries You have laying around in your Junk-Drawer,
You are going to have problems, guaranteed.

Multiple-Cell-Batteries need to be bought NEW, in a "matched-set",
and then permanently kept and used as a "matched-set".
If one Battery of the matched-set develops a noticeable problem,
You should replace the entire set, all at the same time.

The "set" should always be charged, and balanced, and discharged, together, as a set.

There are ways to get around this rule,
but they are expensive and complex,
they also will require continuous attention, and monitoring,
for reasonably reliable performance.

Not following these rules is just begging for trouble.
.
.
.

 

if the cells are discharging faster than they can maintain balance), then i think the cells might not reach fully discharged voltage at the same time.

If the cells are in parallel then, by definition, they will have the same voltage as they discharge, no matter what the load conditions are, as it would be impossible for their two voltages to be different.

 

In parallel, weak cells will provide less current, but voltage remains equal across the cells. Monitoring current through each cell will give a better picture of health.

 

Parallel cells are easier to charge from 5V. Series cells either require a step-up charger (chips are available for USB-C) or a dedicated charger (wall wart, brick). As for discharge, I'd expect that buck conversion (step-down) is more efficient than step-up since currents are lower and there's more voltage available for the MOSFET gates.

 

without any boost system that would be less than 100% efficient.

Unclear. Your scenario without a boost isn't a scenario i'm comparing. My comparison in the question is with a boost.

If one cell was at a lower voltage, current would actually flow into that cell.

Isn't that typically done through shunt resistors? The balance isn't instantaneous, right? It takes a certain amount of time for charge to transfer.

In nicad strings I have seen one series cell destroyed by being reverse charged. Inlarge lead-acid strings I have seen one battery blown up by load current on the series string..

Unclear how that's related.

The "set" should always be charged, and balanced, and discharged, together, as a set.

Sure! But unclear how that answers the original question.

Parallel cells are easier to charge from 5V. Series cells either require a step-up charger

That's a good benefit!

As for discharge, I'd expect that buck conversion (step-down) is more efficient than step-up

My comparison is step-up in both cases.


Comparison

• 10 parallel cells with a 50V boost
• vs 10 series cells with a 50V boost

Which topology is going to perform better under sudden, short-term high current draw? Or no difference in performance?

Which will have a longer discharge cycle?

Which will see more degradation of cells? Shorter lifespan?

More heat?

other differences?

 

14-Cells in series with a Buck-Regulator.

Discharge is only dependent on the Total-Load and Duty-Cycle of the Load.

Life-span is equal.

Cell generated Heat is equal.

A Boost-Converter is very wasteful, roughly ~20% Losses.

In any case,
the Cells must be designed heavy enough to comfortably handle the Load.
Using Cells near their maximum Discharge-Rating will shorten their Life-Expectancy.

Balancing Circuitry MUST BE implemented for Charging.
This is easier for series connected Cells.

What is your expected Peak-Current at ~50-Volts ?
Does the Voltage need to be accurately Regulated ?
Does the Voltage need to be well Filtered ?
Please describe the Load in detail, THE DETAILS MATTER !!!
.
.
.

 

From the cells perspective, there is no difference. Short term high current is a condition imposed on the regulator. If designed for some other steady state draw, you could see horrible efficiencies during that condition, with the higher step up ratio compounding those losses.

 

the Cells must be designed heavy enough to comfortably handle the Load.

You mean current?

I'm guessing the boost is working a lot harder in a parallel circuit. More losses.

Why is Balancing easier for series connected Cells?

What is your expected Peak-Current at ~50-Volts ?

Let's say it's 0.5A. Assume the cells are 1A max.

Does the Voltage need to be accurately Regulated ?
Does the Voltage need to be well Filtered ?

Let's say moderate consumer device, not military/aerospace.
Unclear how this is related. I'm not seeking a practical circuit, just wondering about general pro/cons of parallel vs series cells.

Please describe the Load in detail, THE DETAILS MATTER !!!

Ok, let's say it's an audio amplifier, playing music through a loudspeaker.

Short term high current is a condition imposed on the regulator. If designed for some other steady state draw...

"Other"?

...you could see horrible efficiencies during that condition, with the higher step up ratio compounding those losses.

So, parallel -> 50V boost is much less effiecinet than series -> 50V boost, right?

Thx!

 

The Peak-Current requirements of the Load are a very important consideration.

A Buck-Regulator is usually more efficient than either a Buck-Boost-Regulator, or a Boost-only-Regulator.
The efficiency is rarely in excess of 90%, and usually around ~75 to ~80%,
This is ~20% of your Battery-Power being wasted as HEAT.

Parallel connected Cells need to be electrically isolated from each
other during Charging for maximum life-expectancy of the Cells.
This is not required when Charging Series-Connected Cells, but there must be
"Cell-Balancing" circuitry designed into the Charging-System in any case.

~50-Volts at ~500mA is an unusual combination.
The higher the Voltage requirement,
the easier it is to accidentally over-stress the Cells with excessive Current.
.
.
.

 

the easier it is to accidentally over-stress the Cells with excessive Current.

And i was thinking parallel can deliver higher current. But i guess it's not high current when we're boosting the V so much.

 

Paralleled-Cells can deliver twice as much Current,
but most small Li-Po's can comfortably deliver several Amps without issues,
so a ~500mA Load is not a problem with the Cells connected in series.

"" But i guess it's not high current when we're boosting the V so much. ""
This statement doesn't make any sense.

~500mA would generally not be considered "high-Current" for most common Li-Po Cells
regardless of how many Cells are connected in series to increase the Voltage.
But,
with ~50-Volts available, serious consideration should be given to
including a means of limiting the maximum Current from the Battery.

You could place ~100, ~3.7-Volt Cells, in series, and get ~370-Volts.
This only starts to become a problem when You start demanding
more than around ~1-Amp of Current,
( from smaller sized Cells, really big Cells can deliver hundreds of Amps ).

Or ...........
When You need a higher Amp/Hour-rating,
or need more Average or Peak Current from what are actually under-sized-Cells,
paralleled Batteries of smaller-Cells are sometimes used.
But a much better plan is to increase the Current-capacity of the individual Cells,
rather than using paralleled, under-sized, Cells to get the needed Current-capacity.

Not using paralleled Cells also reduces the complexity of the Charging-Circuitry.
.
.
.